Why your Amazon ROAS is up and your sales aren't
A rising Amazon ROAS with flat revenue may signal that branded search is carrying the result. Before approving more budget, check whether your ads are creating demand or simply claiming credit for it.
If return on ad spend (ROAS) is rising while total Amazon revenue stays flat, check the branded keyword spend. These ads often convert well because shoppers are already looking for the brand. The account may be paying for sales that would have happened anyway.
Meanwhile, category, competitor and benefit-led campaigns have less budget to reach new shoppers. The ROAS report looks healthy, but sales stay where they are.
ROAS measures attribution, not causality
Amazon ROAS is ad-attributed revenue divided by ad spend. It shows how much revenue Amazon connected to the campaign within its attribution rules. It does not show what would have happened without the advertising.
That difference is incrementality: the additional value created because the advertising ran. NielsenIQ describes using ROAS as a substitute for incrementality as a category error. Attribution assigns credit; incrementality asks whether the spend caused an additional outcome.
Flat revenue does not automatically make branded advertising the culprit. Inventory, pricing, promotions and organic performance can also move the numbers. But branded spend is one of the first places to look.
Brand defence can make efficiency look like growth
Branded campaigns reach shoppers searching for your brand or products. They often convert efficiently because awareness and intent exist.
Branded advertising can strengthen visibility, guide shoppers to the right listing and help the brand compete for high-intent searches.
The problem begins when defence crowds out acquisition.
Category, competitor and benefit-led targeting reaches shoppers who have not chosen a brand yet. These campaigns may return a lower ROAS at first, but they create an opportunity to win new demand. Judge them against margin, new-customer contribution and growth, not ROAS alone.
RocketAMS generally recommends keeping branded keyword spend below about 20% of total ad spend. Above that level, brand defence can start taking budget from campaigns designed to reach new shoppers.
Check intent mix and format mix
Two budget splits matter:
Branded versus non-branded intent across Sponsored Products and Sponsored Brands.
Spend and revenue across Sponsored Products, Sponsored Brands and Sponsored Display.
Review both splits. The intent mix shows whether spend is defending the brand or finding new shoppers. The format mix shows where that investment sits across Sponsored Products, Sponsored Brands and Sponsored Display.
Reallocating the existing budget can preserve visibility and give acquisition campaigns room to prove themselves.
What rebalancing can achieve
An Australian brand came to RocketAMS with a 2.25x ROAS and branded search consuming 46% of ad spend.
We kept the budget and redirected investment into category, competitor and benefit-led targeting. Each campaign had one job: defend demand, reach new shoppers and convert active consideration.
Within five months, branded search had fallen to 13% of spend, ROAS had increased to 5.6x and ad-attributed revenue had grown alongside it.
This does not mean every account should target the same splits, but it shows why budget structure should be examined before bids or total spend are changed.
Five numbers to request before approving more budget
Ask your team or agency for:
- Branded and non-branded search spend across Sponsored Products and Sponsored Brands, classified from customer search terms where possible.
- Spend, attributed revenue and ROAS by ad format.
- Total Amazon revenue and TACoS alongside ROAS, using the same window and a comparable prior period.
- New-to-brand performance where available, for Sponsored Brands and Display, alongside acquisition spend and margin.
- Campaigns paused or reduced: their role, spend, profitability, and reasons.
These reports can diagnose budget concentration. They cannot prove causality. Credible incrementality measurement may require controlled testing and stronger analytical methods.
Skai and Stratably’s 2026 US survey found marketers actively measuring incrementality most often reported reduced wasted spend and increased new-customer acquisition. The results are self-reported, but show why incrementality belongs in budget conversations.
What share of ad spend should go to branded search?
The right share depends on brand awareness, category maturity, competition, margin and commercial goals.
If branded search is above 20%* of total ad spend while revenue remains flat, review the budget mix before increasing spend.
Look at the split before the bids
If ROAS is rising while total revenue remains flat, ask where the budget is working before asking it to work harder.
RocketAMS can review your historical Sponsored Products, Sponsored Brands and Sponsored Display performance to show how much spend supports defence, where acquisition has been restricted and which campaigns may suit reallocation.
Want to see where the growth opportunities are? Request an Amazon Ads spend review.
Frequently asked questions
Yes. ROAS shows how much ad-attributed revenue was recorded for each dollar spent. It doesn't show whether the advertising created an additional sale. A higher ROAS may come from shifting more budget into branded searches, where shoppers already know what they want and are more likely to convert. The report improves even if total revenue and customer acquisition remain flat.
Review ROAS alongside total Amazon revenue, TACoS, branded spend share, non-branded sales and new-to-brand performance where available. Also check changes in price, promotions, inventory and organic visibility during the same period. Amazon Ads notes that ROAS may not be the most informative measure when the goal is awareness or repeat purchases. NielsenIQ makes the wider distinction: attribution gives advertising credit for a sale, while incrementality asks whether that sale would've happened without the spend.
There's no universal percentage, and Amazon doesn't publish a recommended branded-spend share. RocketAMS treats branded search above roughly 20% of total ad spend, alongside flat revenue, as a reason to investigate the account. It's a diagnostic threshold rather than a target every brand should follow.
The appropriate share depends on brand awareness, competitor activity, reseller presence, category maturity, margins and the account’s growth objective. A well-known brand facing heavy competition on its own terms may need more defensive investment. A growing brand with limited category reach may need a larger share directed towards non-branded acquisition.
Classify spend using actual customer search terms across Sponsored Products and Sponsored Brands. Campaign names alone can mislead because broad or automatic campaigns may capture branded searches. If the share appears high, reduce it gradually and monitor total branded sales, organic sales, TACoS and non-branded growth before settling on a new mix.
Not automatically. Branded campaigns can help the brand maintain prominent placements, direct shoppers towards priority products and reduce the opportunity for competitors or resellers to intercept high-intent searches. They can also support launches, promotions and variation control.
Their value still needs to be tested. A shopper searching for the exact brand may have purchased from the organic result anyway, so attributed sales can overstate the additional demand created. Keep branded targeting in clearly separated campaigns, then test controlled changes to bids or budgets rather than switching everything off at once. Track total branded sales across paid and organic traffic, Featured Offer performance, competitor visibility, TACoS and overall revenue during the test.
Retain the spend that provides commercially useful defence. Reallocate the portion that mainly claims credit for existing demand, particularly when category and benefit-led campaigns are being restricted by budget.